Scoping Review: A Needs Based Assessment and Epidemiological Community-Based Survey of Ex-Service Personnel and their Families in Scotland

SECTION 6: Summary, Conclusions and Recommendations

6.1 PREFACE

The purpose of this section is to conclude by summarising the key evidence that emerged from the consultation process and the review of the eminence- and evidence-based literature and to consider the case for undertaking a methodologically robust needs-based assessment and population-based survey of Veterans and their families in Scotland. In so doing, the extent to which similar studies conducted elsewhere might offer a valid blueprint for such a study are explored and potential models of funding and delivery are recommended.

6.2 Summary of findings: Consultation process

The following key issues and suggestions emerged from the interviews conducted with stakeholders and Veterans.

6.2.1 Pre-Service Factors

• Pre-enlistment vulnerabilities play a major factor in determining those who fare least well on leaving Service and who are at an increased risk of social exclusion.

6.2.2 Service-Related Factors

• Paradoxically, whilst military life offers a number of positive benefits in terms of providing structure, security, a sense of identity and an esprit de corps, these benefits may exact a penalty on demobilisation and the transition to civilian life including the risk of fostering a "dependency culture".
• Those who enlist in the Army and of a lower rank experience the greatest problems in adjusting to civilian life.
• The prevailing culture/ethos of "machismo" and "toughness" does not readily accommodate the need to report mental health problems and seek help. It also was also suggested as being a major reason why the heavy use of alcohol is a common feature of military life.
• Stigma is a prevailing issue in relation to mental health problems and help seeking.
• Support for partners of Service personnel is lacking, particularly for Reservists.
• Significant advances in military prehospital care and trauma management have generated a new major challenge in terms of the long term physical and emotional adjustment for severely injured personnel.
• Efforts by the Services to facilitate the transition from military to civilian life are inadequate, particularly for those who are most vulnerable to transition difficulties such as Early Service Leavers and those discharged on medical grounds. The primary problem is not so much the volume of information provided but the manner in which it is presented and/ or reported.

6.2.3 Post-Service Factors

• Differences in need exist between older and younger Veterans.
• A number of barriers to provision of mental health care for Veterans exist including:
  • attitudes and behaviours of the individual (e.g., fear of stigma, denial, excessive alcohol use);
  • attitudes and behaviour of others (e.g., employers, healthcare professionals, GPs), and
  • delivery of mental health services for Veterans (e.g., lack of specialist trauma services, conflict of interest between voluntary and statutory agencies, lack of integration among agencies).
• For some, the transition from military to civilian life includes a number of social and environmental problems in terms of:
  • employment (e.g., lack of experience in applying for a job)
  • housing (e.g., lack of sympathy from LAs)
  • finance (e.g., inability to manage financial affairs)
  • support for partners (e.g., lack of recognition for supportive role to ex-Service partner)
  • personal (e.g., perceived loss of special and valued identity)
  • social exclusion (e.g., due to alcohol/alcohol misuse, unemployment, debt, homelessness, mental ill health)

6.3 Summary of findings: Review of eminence- and evidence-based literature

• Overall, the evidence suggests that military life is beneficial for the majority who serve and that their transition to civilian life is successful. However, there is a significant minority who fare badly, particularly Early Service leavers (i.e., those who have completed less than four years of Service) and those who have served time at the MCTC prior to discharge. Both of these groups are more likely to have had a previous history of childhood anti-social behaviour than their Serving counterparts.
• Military personnel with mental health problems are more likely to leave Service early and have an increased risk of ongoing social exclusion and ongoing ill-health.
• Pre-enlistment factors known to affect the risk of adverse health and wellbeing outcomes include:
  • childhood traumatic experiences
  • socio-economic adversity
  • previous psychiatric history
  • personality
  • coping style
• Single males, of lower rank, with lower educational status and who have served in the Army are most at risk of pre-service vulnerability factors experienced in childhood. To what extent, however, this association would be significantly different from a similar age-matched group in the general population has yet to be established. Moreover, it is not known to what extent these findings would generalise to women.
• Recent KCMHR studies of Veterans report a prevalence rate of 4% for "probable" PTSD, 19.7% for symptoms of common mental health problems and 13% for alcohol misuse.
• Reports from the charitable sector suggest that presentation with combat-related mental health problems can be as long as 14 years post-discharge. There is however no evidence to suggest that ex-Service personnel are no more or less likely to seek help than people who have never served.
• Compared with their Regular counterparts, Reserve UK Armed Forces personnel (particularly medical Reservists) have an increased risk of experiencing mental health problems as a consequence of deployment to Iraq and Afghanistan. However, the evidence suggests that this finding is more likely due to family issues prior to deployment, support to families during deployment, and experiences of home coming than events in theatre.
• Relative to the general population, both Serving and ex-Service personnel report higher levels of alcohol consumption (but only in younger age groups). Alcohol misuse has also been identified as a problem affecting Service women.
• The overall rate of suicide is no greater among UK ex-Service personnel than in the general population. However, for men aged 24 years and less who have left the UK Armed Forces the risk of suicide is approximately two to three times higher than that of the same age group in both the general and serving populations.

6.4 RESEARCH RECOMMENDATIONS

• Underpinning an evidence-based approach to Veteran care requires the necessary epidemiological data and needs assessment to quantify the clinical and economic burden of ill-health and characterise the excess morbidity reported as well as to identify those factors which conduce to a positive transition from the UK Armed Forces to civilian life including resilience, recovery, and post-traumatic growth (Sundin et al, 2010). Whilst a considerable body of literature exists on the health and wellbeing outcomes in respect of serving military personnel, few studies have been specifically dedicated to those who have left the UK Armed Forces.
• Most of the literature relating to ex-Service personnel in this domain derives from the US. The extent to which the findings from such studies are generalisable to ex-Service personnel in the UK is limited by virtue of two key differences that exist between these two nations, viz, the Vietnam War experience and the subsequent establishment of the US Veterans Administration designed to provide bespoke medical and psychiatric treatment for US Veterans (Fear et al, 2009). Within the UK, most of the academic research of note has been conducted by the KCMHR and is predominantly based on their military cohort data. However, to inform service provision and the development of appropriate interventions it is imperative to establish the extent to which the Veteran population differ from those in the general population. Whilst there have been two studies in the UK that have adopted this approach, one in England (Woodhead et al, 2011a; 2011b), and one in Wales (http://medicine.cf.ac.uk/primary-care-public-health/research/healthy-places/completed-projects/Veterans-needs-assessment/), comparisons between the two are limited due to variations in the design, sampling strategies, and method of diagnostic assessments used.
• There is a paucity of robust epidemiological data about the health and wellbeing, views, expectations, and needs of ex-Service personnel and their families in Scotland, and how these compare with the general population. The current evidence-base is therefore not sufficient to design specific health strategies or develop new services-within or outside the conventional health care system-particularly in order to reach non-treatment seeking ex-Service personnel who suffer from mental health problems and are at risk of the downward social spiral that can lead to breakdown in relationships, homelessness, debt, and encounters with the criminal justice system. A key related factor that requires further investigation is the role of alcohol with regards to increasing the risk of social exclusion.
• Historically the profile of the UK Armed Forces has not been representative of the general population. To deliver operational capability, the UK Armed Forces employ people who meet specific standards of health and fitness and certain nationality criteria. The ratio of men to women in the UK Armed Forces is substantially higher than would be the case in most other occupations; a circumstance which in part derives from the fact that the military are unique in having to conduct operations on behalf of the nation, which if necessary involves combat. Although the number of Service women has gradually increased in line with the implementation of equal opportunities policies by the MoD, there is a paucity of research about their specific health and wellbeing needs, particularly post-Service.
• Given the evidence that military service has a positive impact for the majority of UK Armed Forces personnel, it is important to study not just those who may be damaged by service but also those who gain by it. Thus, whilst a greater understanding of pre-enlistment vulnerability risk factors is required, it is also necessary to understand what factors facilitate and enable those with the same vulnerabilities to fare well by using appropriate comparison groups. Such an approach would enable the UK Armed Forces to identify what could be done to improve the life chances of young people with pre-enlistment vulnerabilities in preparation for leaving military service.
• Many regiments have historical regional areas of recruitment, which for traditional forces include areas of economic and social deprivation (Fossey, 2010). Although evidence suggests a link between areas of high social deprivation and offending behaviour, it is not known for example whether recruitment into the UK Armed Forces from such areas has a positive or negative impact on the risk of subsequent offending and incarceration of those particular recruits on leaving military service and to what extent regional variation exists.
• Relatively little is known about the health and wellbeing outcomes in the longer term for ex-Service personnel due to the paucity of large prospective longitudinal studies. Well-designed cohort studies offer a number of advantages in respect of identifying the longer term implications of military service (Fear et al, 2009).
• Epidemiological data are lacking on specific groups who are at risk of health and wellbeing problems such as the wounded and disabled and those exposed to high intensity combat (Deahl et al, 2011). For example, in respect of the increasing numbers of injured combat troops returning from Iraq and Afghanistan with blast related polytrauma (Sayer et al, 2008) and associated traumatic brain injury [TBI] (French & Parkinson, 2008). Relatedly, there is a particular need to understand the longitudinal course of post-traumatic reactions and to assess the factors which may affect the reporting of symptoms at follow-up such as persisting physical problems, anticipation of returning home and starting work, concern over healthcare, and pending disability. Relatedly, whilst the acute clinical management of polytrauma has been well described, the post-acute clinical management of such cases in rehabilitation and primary care settings has not been adequately addressed in the literature. Very little is known about the impact of military and ex-Service life on partners and families. Research that takes into consideration military personnel with different family configurations (e.g., dual-deployed parents and single parents) is urgently required. In 2009, the ACTR was commissioned by the Headley Court Trust to undertake a 3-year longitudinal study to identify factors which conduce to and prevent rehabilitation and adjustment in military personnel and their partners following combat-related injury (Forbes et al, 2011a; 2011b). The Final Report for that study is due to be submitted to the Headley Court Trust in January 2013.
• An increase in the deployment of Reservists in the UK Armed Forces and concerns about their health outcomes following deployment highlights the need for a greater understanding of the impact of the additional challenges facing families of Reservists compared to the family members of Regular personnel.
• To date there has been no health economic evaluation of the financial impact of providing health and welfare support services to ex-Service personnel. McCrone et al (2003) highlight this gap in the literature with specific reference to the health economic considerations relating to service interventions for ex-Service personnel who are diagnosed with PTSD.
• There remain gaps in the evidence with regards to determining the effectiveness of health and welfare interventions in relation to:
  • establishing whether recommended treatments are best delivered in specialist Veterans' mental health facilities or in the mainstream NHS (Fossey, 2010);
  • identifying factors that facilitate or inhibit help-seeking (e.g., due to a fear of stigmatisation) and to establish whether differences do indeed exist between ex-Service personnel and the general population;
  • identifying factors that facilitate resettlement, in particular for those who are at greatest risk of experiencing problems in making a transition from military to civilian life (e.g., Early Service Leavers);
  • promoting social inclusion (e.g., by enhancing employment opportunities, and reducing the number, experiences and outcomes of homeless ex-Service personnel), and
  • health economic considerations.

6.5 needs-based assessment and population-based survey of Veterans and their families in Scotland

• The principal aim of this scoping review was to identify to what extent a population-based survey is required to inform the national commitment to meeting the health and wellbeing needs of ex-Service personnel and their families in Scotland. The outcome of this comprehensive review has confirmed that there is a paucity of robust epidemiological data about the health and wellbeing, views, expectations, and needs of ex-Service personnel and their families in Scotland, and how these compare with the general population. On this basis, there are sufficient grounds to warrant consideration of the case for undertaking a methodologically robust needs-based assessment and population-based survey of Veterans and their families in Scotland. The outcome of such an endeavour is that it would provide an important first step to informing the scale and nature of health and welfare provision (Busuttil, 2010) and facilitating the development of a "flexible, needs-led, high quality standard of care" (Deahl et al, 2011).
• Such an approach is in accordance with the promotion of a well-being and recovery-based mental health community-based service model produced by the former Scottish Executive's Mental Health Division (DfMH, 2006), which combines a population-based approach to social inclusion to prevent mental illness and inequalities in mental health and highlight the link between mental and physical health. The guidance in DfMH is based on evidence of what works in terms of achieving better outcomes for individuals through using appropriate services that meet their needs. It promotes a functional model of service design and requires local partners to ensure service effectiveness and achieve high standards in response to local needs.
• Presented below are the key issues that need to be considered in order to determine how best to undertake such a study, the approach for which has been informed by that of Anderson et al (2008b) in undertaking a Scoping Study to identify the need for a Scottish Longitudinal Study of Ageing.

6.5.1 Identification of an Appropriate Comparison Group

• Identification of an appropriate general population comparison group is potentially problematic due to the socio-demographic profile of the UK Armed Forces (Fear et al, 2009). One option would be to identify specific index and comparison groups against that could be meaningfully contrasted with other occupational groups to control for the "healthy worker" effect such as first responders (e.g., fire-fighters, police, and paramedics).

6.5.2 Data Protection and Ethical Issues

• The UK's Data Protection Act (1998) has had a substantial impact on health-related research. At present the 1998 Act allows medical data to be used for any medical research purposes without the need for consent of individuals. Changes in the implementation of UK data protection law has placed major bureaucratic obstacles for epidemiologists to overcome in seeking to trace individuals who have not yet consented to participate in research (Iversen et al, 2006). Consequently most codes of ethical conduct recommend informed consent for any medical research whether it involves direct contact with participants or access to their records. To obtain consent however incurs participation bias and inevitably reduces the response rate, in particular when using a sampling frame that prevents researchers immediate access to contact details to make a direct approach to potential participants. Those who are most likely to respond are those for whom the study has the greatest salience. Consequently the sample size is likely to become too small and lack sufficient statistical power to reach generalisable conclusions (Iversen et al, 2006).
• A pertinent illustration of the issues raised above is the outcome of a pilot conducted by Klein et al (2004) to test out the feasibility of conducting an epidemiological community-based clinical, psychological and economic survey of post-traumatic stress disorder and trauma-related psychopathology in Grampian. To maximise the efficiency of the pilot study in capturing variation by socio-economic status and area of residence (urban vs rural), a random sample of 500 adults (18 years and above) registered for primary care, and stratified according to census-based indicators of deprivation and degree of urbanisation of residence, were selected by means of the Community Health Index (CHI); a computerised database of all patients registered with general practitioners in Scotland. Five stages were required before it was possible to conduct the telephone screening interviews, as follows.
  • Stage 1: Prior to commencing with recruitment, full ethical approval was obtained from the Grampian Research Ethics Committee, and permission was obtained from the Director of Public Health for access to the CHI.
  • Stage 2: Randomisation and identification of potential participants was conducted by the Data Manager (Data Management Team, College for Life Sciences and Medicine, University of Aberdeen).
  • Stage 3: In accordance with ethical procedures governing the use of the CHI, a letter was sent under the aegis of the Director of Public Health to all the general practitioners, whose patients had been identified, to obtain approval for the inclusion of their patients in the study. A total of 27 (5.4%) patients were identified by this means as being ineligible to participate on clinical grounds. Thus reducing the number from 500 to 473.
  • Stage 4: As the research team were not permitted to know the identity of the 473 (94.6%) eligible individuals until consent to participate had been obtained, a letter was sent by the Director of Public Health inviting them to take part in the study and requesting them to complete a consent form and a contact details form to be returned directly to the research team.
  • Stage 5: Once permission was finally granted for the research team to contact the 473 individuals who had consented to participate, a response was received from 133 (28.1%), 29 of whom subsequently declined to participate. A total of 104 telephone screening interviews were completed.
• In an endeavour to promote health care research, the new General Medical Council (GMC) Guidelines on Confidentiality (September 2009) have indicated a more proportionate approach to using health care data (Fear et al, 2009).

6.5.3 Tracing Issues

• For reasons already described above, a high participation rate is recognised as a key priority for any research study. However, health-related research involving ex-Service personnel presents a considerable challenge in this regard since the group least likely to return questionnaires in the general population are typically young single males (Tolonon et al, 2006) who represent a significant proportion of this target population. The problem is further compounded by virtue of the fact that on leaving military service it becomes increasingly difficult to trace individuals (Hotopf & Wessely, 2005) as obtaining a valid address becomes increasingly difficult with the passage of time, in particular with regards to those service leavers who are peripatetic due to frequent changes in employment and accommodation. Evidence suggests that such factors are associated with mental health problems and increase the risk of social exclusion. Whilst tracing these individuals is notoriously difficult, a recent article by Fear et al (2010) provides a helpful account of the methods used by KCMHR when conducting a telephone survey to identify the prevalence of mental health, health service usage, and stigma within serving and ex-Service personnel. Methods to trace participants included the:
  • military database that contains last known address;
  • NHS address registry (but this is contingent on the individual being registered with a GP), and
  • next of kin contact details.

All contact details were subsequently cross-referenced against the electoral roll to determine if resident at the specified address.

• Fear et al (2010) concluded that the use of "multiple simultaneous" tracing methods and tailoring the approach in accordance with the target population helps to increase rapport with participants and encourage compliance.
• Efforts have been made to improve the recording of data relating to identifying those who have served in the UK Armed Forces. For example, as a result of communications between DASA and ONS, a question to this effect has been included in the 2011 Census.

6.5.4 Regional Variation

• Within Scotland there are distinct regional differences in terms of poor health, with different rates depending on deprivation, remoteness and the urban/rural setting. In addition, there are distinct regions that are typically associated with the military. This has implications in respect of the sampling strategy to ensure a nationally representative sample.

6.5.5 Sampling Frame

• The sampling frame operationally defines the target population from which the sample is drawn and to which the sample data will be generalized. most familiar type of probability sample is the simple random sample, for which all elements in the sampling frame have an equal chance of selection, and sampling is done in a single stage with each element selected independently (rather than, for example, in clusters).
• Somewhat more common than simple random samples are systematic samples, which are drawn by starting at a randomly selected element in the sampling frame and then taking every nth element (e.g., starting at a random location in a telephone book and then taking every 100th name).
• In yet another approach, cluster sampling, a researcher selects the sample in stages, first selecting groups of elements, or clusters (e.g., census tracts), and then selecting individual elements from each cluster (e.g., randomly or by systematic sampling).
• Three main sampling options have been identified as being available for a study of this type.

(i) Bolt on to an existing national survey.
(ii) Screen addresses to identify eligible households/ individuals.
(iii) Identify a sample from an individual-level sampling frame or database.

• The particular advantages and disadvantages of each of these three sampling options is summarised below.

(i) Bolt on to an existing national survey

• Based on the broad range of health and welfare outcomes identified in this scoping review, four Scottish national data sources were considered as potential sampling frames on the grounds that they have been designed and implemented exclusively in Scotland to address specific Scottish issues and contexts alongside the more general demographic information needed to cross classify them.
  • Scottish Longitudinal Survey (SLS)^[181] - The SLS is a large scale linkage study that links data from the Census with data provided by various administrative and statistical sources (including health records). The 1991 Census was used to identify approximately 274,000 SLS members and information from these individuals have been linked to other datasets including the 2001 Census, vital events and health information. It has been widely used to study health variations over time and has the benefits of low attrition (because the data collected are either required by law or as part of a standard administrative data) and high linkage with events.
  • Scottish Health Survey (SHeS)^[182] - The SHeS provides a detailed picture of the health of the Scottish population in private households and is designed to make a major contribution to the monitoring of health in Scotland. It is regarded as being essential for the Scottish Government's forward planning, for identifying gaps in health services provision and for identifying which groups are at particular risk of future ill-health. The specific aims of the SHeS are to:
    • estimate the prevalence of particular health conditions in Scotland;
    • estimate the prevalence of certain risk factors associated with these health conditions and to document the pattern of related health behaviours;
    • identify differences between regions and between subgroups of the population in the extent of their having these particular health conditions or risk factors, and to make comparisons with other national statistics for Scotland and England;
    • monitor trends in the population's health over time, and
    • make a major contribution to monitoring progress towards health targets.

From 2012-2015 the survey has been designed to produce an achieved sample size of around 4,000 adults and 1,800 children per year.

• Scottish Crime and Justice Survey (SCJS)^[183] - The SCJS is an annual social survey that involves interviewing a randomly selected adult in 12,000 households across Scotland identified by means of the Postal Address File. The principal aims of the SCJS are to:
  • generate reliable statistics on peoples experience of crime, including services provided to victims of crime;
  • evaluate the varying risk of crime for different groups of people in the population;
  • examine temporal trends in the level and nature of crime in Scotland;
  • obtain information about people's experiences of, and attitudes on a wide range of crime and justice-related matters;
  • provide an alternative and complementary measure of crime to the police recorded crime statistics, which provide statistics on crimes and offences recorded and cleared up by the eight Scottish police forces.

Findings from the SCJS are used by policy makers across the public sector in Scotland to help understand the nature of crime in Scotland, target resources and monitor the impact of initiatives to target crime.

• Scottish Household Survey (SHS)^[184] - The SHS is designed to provide accurate, up-to-date information about the characteristics, attitudes and behaviour of Scottish households and individuals on a range of issues. It is a "continuous cross-sectional survey", each complete sample being covered in the course of two years to amount to 31,000 households. The sample is being drawn from the small user file of the Postcode Address File.
• Whilst there are a number of potential advantages to bolting on to an existing national survey (particularly in terms of reducing cost) discussion with those responsible for overseeing these national surveys confirmed for a variety of reasons it would not be a viable option.
• Although the Adult Psychiatric Morbidity Survey (APMS)^[185], which is another nationally representative survey, was used to generate a community-based sample to compare the health of conscripted national service Veterans with population controls (Woodhead et al, 2010b), it would not be suitable for the proposed Scottish survey because:
  • it is based on a community-dwelling sample of adults (aged 16 years and above) in England only, and
  • the sample size would be insufficient to draw reliable conclusions about younger Veterans given the majority were aged over 65 years.
• Discussion also took place with Professor Simon Wessely (Director, KCMHR) and Dr Nicola Fear (Reader in Epidemiology, ACDMH) to explore the possibility of identifying a sample of ex-Service personnel from the KCMHR 2003 cohort in light of plans to re-contact 12,000 Service personnel of whom 4,000 had since left Service and 400 were identified as residing in Scotland. However, whilst this approach would potentially provide a representative sample of ex-Service personnel (given that it is based on a random selection), the question remains as to how best to identify a suitable comparison group that would be representative of the Scottish general population.
• An alternative approach was recently reported that used the KCMHR 2003 cohort to provide evidence to inform service provision for military ex-Service personnel residing in Wales (http://medicine.cf.ac.uk/primary-care-public-health/research/healthy-places/completed-projects/Veterans-needs-assessment/). It was funded by the Welsh Assembly Government (WAG) to develop a service specification for a Wales-wide service for Veterans and led by Professor Jonathan Bisson (Department of Psychological Medicine and Neurology, Cardiff University). The study was a cross-sectional survey to assess the needs of military Veterans in Wales and involved a telephone study of a sample recruited from the following three sources.
  • The KCMHR randomly selected cohort of all 261 Veterans who were living in Wales.
  • A random sample of 150 Veterans out of the total of 262 living in Wales
    and in contact with Combat Stress.
  • A random sample of 150 Veterans out of the total of 12,017 living in Wales who had made contact with the SPVA.

(ii) Screen addresses to identify eligible households/individuals

• The main advantage of a new sample based on screening of households from the Postcode Address File is that it has no "inherited" response bias and it offers more flexibility with regards to geographic clustering as data collection is not tied to other surveys. Its main disadvantage it that it is typically considerably far more expensive to conduct and can be unpopular with interviewers, who prefer to spend their time interviewing rather than screening and, as such, adversely affect response rates.

(iii) Identify a sample from an individual-level sampling frame or database

• The ideal option would be to draw a fresh sample from an individual-level database containing former military service as an identifier variable. In Scotland, the National Health Service Central Register (NHSCR) offers potential opportunities as a sampling frame for this approach. It maintains records of all NHS patients and birth records of those not registered with a GP (unlike the Community Health Index which is a listing of everyone who has been or is registered with a Scottish General Practice) including more recently those who have joined the UK Armed Forces. As an electronic administrative database it enables the:
  • capturing of data on migration to identify exits and re-entry from one Health Board to another and from NHSScotland to retain contact with individuals who relocate within the UK by virtue of co-operation with NHS Information Centre (NHSIC) and Medical Research Information Service (MRIS), Southport;
  • flagging of the date of death of consenting participants (a key factor given the shifting demographic profile of an increasingly older population), and
  • availability of record linkage techniques to match self-report data with medical records held by the ISD National Health Service (NHS) Scotland and facilitate the economic analysis.
• Use of the NHSCR for a military study has been piloted in Scotland. In principle, the NHSCR would offer an ideal sample frame and would lend itself well to the data linkage aspects of the study. However, there would be potential problems about achieving a sufficiently large sample size for the same reasons described in 6.5.2 above with regards to the use of the Community Health Index in the Grampian pilot (Klein et al, 2004). For this reason, the costs involved would be considerably more than those for the other two approaches identified above.
• A pertinent example of this type of approach is an epidemiological study of trauma, health and conflict conducted in Northern Ireland (Ferry et al, 2008). The rationale for the study was based on the fact that because Northern Ireland has experienced 30-40 years of civil conflict in its recent history (often termed the "Troubles") there was a need to identify the traumatic impact on health and trauma-related disorders. It comprised two parts. The first was based on the Northern Ireland Study of Health and Stress (NISHS), which represents one of the largest ever population based studies of mental health undertaken in Northern Ireland and is part of the World Mental Health (WMH) surveys conducted in over 28 countries worldwide under the aegis of the WHO. As is the case for all WMH surveys, the sampling frame is based on the generation of random numbers to select a multistage household probability sample using clusters in order to link in with Government figures and Census output areas. At the time of the study a total of 3,100 individuals had been interviewed face-to-face using the WMH-Composite Diagnostic Interview (WMH-CIDI; WHO, 1997); a comprehensive, fully structured interview designed for use by trained lay interviewers for the assessment of mental, behavioural, and substance use disorders according to the definitions and criteria of ICD-10 (WHO, 1992) and DSM-IV (APA, 1994). A number of screening questions are asked to determine the direction of the interview. To achieve that number of interviews required the identification of 6,000 in the sampling frame and cost in the region of £1 million, which was spent on postgraduate awards although the majority of costs were attributed to undertaking the interviews as the data collection for the NISHS was outsourced given that it required 40 trained interviews. The second part of the study used qualitative methods to find out more about the experiences of individuals who had been traumatised by their experience of civil conflict in Northern Ireland. Data was collected over a period of three to four years.
• An interview with Professor Brendan Bunting (Professor of Psychology, Psychology Research Institute) as the lead for the Northern Ireland study revealed that the key issues apart from the cost were as follows. First, although the age for eligibility to participate was 16 years and above, he recommended increasing it to 18 years and above. Second, because of legislation in Northern Ireland dating back to 1968 which required the passing on of information regarding knowledge of any criminal act, this required the interviewer to state as part of the informed consent process that any such knowledge would have to be reported to the authorities. Third, there was a problem in terms of determining how best to deal with mental health problems identified as part of the assessment process. Two psychiatric social workers were subsequently employed on a standby list to deal with any participants in need of help. Fourth, each participant was paid £10 per interview. Fifth, in addition to providing a greater understanding of some of the experiences reported in the first part of the data collection, the qualitative interviews were also of benefit in terms of enhancing the face validity by providing a context and bringing to life the personal experiences of participants.

6.6 DESIGN AND FUNDING Recommendations

• The possible options for the design of a population based survey to inform policy development in meeting the health and wellbeing needs of ex-Service personnel and their families in Scotland need to take into consideration the following challenges.
  • Identification and tracking of Scottish Veterans and their families to permit a representative sampling frame and strategy.
  • Adequate sample size(s) to ensure sufficient power to detect statistically significant differences thereby maximising the generalisability of findings.
  • Recruitment and retention mechanisms to maximise compliance rates and minimise loss at follow-up thereby enhancing the representativeness of the sample(s).
  • Reliable and valid assessments.
  • Identification of index and comparison groups against which different sub-samples of the Veteran population (e.g., those who are at risk of social exclusion) could be meaningfully contrasted.
  • Data protection and ethical requirements (including identifying mechanisms to ensure the safety of research personnel involved in the recruitment and assessment of participants).
  • Costing requirements associated with possible approaches according to:
    • a realistic and feasible timeframe, and
    • a realistic estimation of the resources required (i.e., research personnel, travel, subsistence, equipment, consumables, consultancy, data entry and analysis).
• The design should be ambitious in terms of sample size to allow for detailed sub-group analyse and to take into account the potential effects of attrition due to the nature of the target population.
• Commitment should be sought for funding for at least 3 years but preferably longer to ensure that the long term value of the resource is fully realised.
• A mixed methods approach should be used to allow for the in-depth understanding of key issues (e.g., barriers to help seeking).
• The design must be founded on the rigorous application of population-based research methods to:
  • assess a broad range of health and wellbeing needs;
  • accurately identify those individuals who require health and social care services;
  • reliably evaluate the use and perceived effectiveness of health and social support and clinical care currently available, and
  • identify factors conducive to recovery, wellbeing, and psychosocial adjustment.
• Given the likely scale of a population based survey in Scotland and the range of interests it is likely to serve, a form of collaborative funding is recommended. One approach would be to seek matched funding from the Scottish Government for resources elsewhere. Other potential sources of funding include the Economic and Social Research Council (ESRC), the Chief Scientist Office (CSO), and the Welcome Trust. This approach proved successful in a recently completed collaborative study involving the KCMHR, the ACTR and the National Centre for Social Research (NatCen), which was funded by the ESRC. The principal aim of the study was to investigate the attitudes and perceptions to the Iraq and Afghanistan missions amongst the British public using the 2011 British Social Attitudes (BSA)^[186] survey as a platform for a module of 40 questions (Gribble et al, 2012). The concept for this study originally derived from one of the findings from this Scoping Review, viz, the potential for stigma experienced by ex-Service personnel to affect their ability to adjust successfully to civilian life in terms of employment, housing, and health care.